Roda group of Temples

Roda group of Temples-
The Pioneering Monuments of
the Architecture Heritage of Gujarat

-Jayaram Poduval,
Published by:Gujarat Tourism, 2005



There is a thin film of mystery shrouded around the temples of our concern. The mystification starts from the name itself. The site has been known as `Roda' to the historians and art historians though the local population is barely aware of this. The site is situated between the village of Raisingpura and Khed Chandarani, 8 Kms from Himmat Nagar in north Gujarat. Curiously, the writings on Roda never mention these villages. Nor do they mention about the derivation or etymology of the name. UP.Shah mentions thus, "The shrines are called Roda Temples after the hamlet of the name in the vicinity, but the name Roda (In Gujarati, literally meaning brick-bats), is not the original name of the spot, which latter seems to have forgotten now, the place being called Roda on account of the extensive ruins containing brick-bats, and an old habitation site situated close to this village."[1]

May be the first explorer at Roda was informed about this ruined stone blocks by the local population, for whom this site was nothing more than that. Today the devotees from the nearby village do not even glance at the cardinal shrines at the site as they walk past them to the Matrika images outside a shrine in the Kunda. Did they forget the main shrines and the principal deities in them? The neglect to the shrines endured in the works of pioneer scholars who traveled around Western India and listed the monuments of archaeological and historical potential. The early chroniclers like James Tod and archaeologists like R.D.Bhandarkar are silent about the site in our purview. The monumental volume of James Burgess and Henri Cousens never mentioned Roda Group of Temples. Surprisingly Sankhalia in his in depth study of Gujarat Archaeology fail to refer the Roda group.


The studies on Roda was initiated by PA Inamdar [1926] followed by UP Shah and MA Dhaky in the 60’s. Since then barring few articles by Dhaky, Roda has never come into the attention of the scholars or tourists. The sculptures at Baroda Museum & Picture Gallery and other collections in Gujarat definitely points to an intense temple building activity in the region prior to the Solanki period. The activity centered in the present Banaskantha and Sabarkantha districts of Gujarat State, bordering Rajasthan on the north and Madhya Pradesh on the east. For example many sculptures at the Baroda Museum including the ones lying in the storeroom could not have been from the existing temples at Roda. The dilapidation seems to be a continuous process at Roda as Temple No. 4 reduced to a mere plinth between the visits of Inamdar in 1926 and that of UP Shah in 1960. Apart from the seven temples, which form the core of the present study, there are hardly any evidences of the existence of any structure at the site. Moreover the style of some of the sculptures appears to be post-dating the extant structures and thus pointing towards an enduring activity at the site. If one accept this hypothesis the present site itself can present a continuous activity of temple building like that of Pattadakkal [Karnataka] and Mahabalipuram [Tamil Nadu]. The centers like Roda, which deprived of its licit place in the history of Indian architecture, need to be placed in the macro chronology through proper studies, as they set the trend of the regional idiom of art and pave the way for magnificent structures of Gujarat as the Sun Temple at Modhera and Navalakha Temple at Sejakpur.

It was Dr.U.P.Shah who initiated the study on the Roda group of temples, through his book "The Sculptures from Shamalaji and Roda". In this, he introduces a pioneer scholar who visited Roda way back in 1926. P.A.Inamdar, as mentioned by Dr.Shah, visited the sites and provided ample data on the structures and the ruins around. Curiously, the only photographic documentation of the Temple no.4 comes from Inamdar, in 1926 in his book titled "Some Archeological Finds in the Idar State". Inamdar's description of Roda is synoptic and a confession of the lack of methodology. He, as stated in the preface, is only concerned about reporting the historical monuments of the region. "In my tour of inspection of the {Idar} state schools, I was struck with the beautiful scenery of the state, which is studded with hills and is full of forests, fertile valleys watered by great rivers, lakes, and last but not least, old monuments in form of beautiful temples, idols of exquisite beauty and anatomical perfection, tombs of ancient heroes and beautiful step-wells. My Interest in these, especially the monuments, grew more and more as I found these archaeological interest and so with a view to preserve these and to prevent them being further damaged ...[sic]"[2]

Nevertheless, this plea to the authorities did not seem to have any effect as we know at least in case of Roda the structures were well on the way to decay and disappearance. In the portion on Roda, Inamdar notices few structures and few mounds. The few temples as noted by Inamdar was dedicated Shiva, Vishnu, Goddess and Surya which can be identified with the existing temples of No.VII [Shiva], No.V [Vishnu], No.II [Goddess] though the Surya Temple seems to have disappeared. The evidence for the existence of the Surya temple can be endorsed with the help of the image of Surya, now in Baroda Museum. The notion, which one gets reading Inamdar, is of a great archaeological site spotted with "beautiful" monuments. "It appears”, writes Inamdar, “some prosperous city renowned as a place of Hindu-pilgrimage, existed here in the old times as is evident from the large number of temples which are partly destroyed and in many cases exist as mounds only"[3]

Naturally, a question can be asked why one considers Shah as the person initiated the study on Roda. It was for the reason that it was Shah who concentrated on Roda and realized it’s potential as monument of merit and significant in the study of Western Indian temple architecture. The concluding words of Shah on the chronology of Roda are evident of his admiration and understanding of the importance of these structures. Shah states, " This {the comparative study of Roda along with the Saindhava and Maitraka temples of Gujarat} is likely to throw a new light on the history of temple architecture in India"[4]. Shah's work however urged at least some scholars, mainly from Gujarat to take notice of the monuments of Roda. Soon after the publication of the book by Shah, M.A.Dhaky in his article "The Chronology of the Solanki Temples of Gujarat" marked the place of Roda as the fountainhead of Solanki style of architecture in Gujarat. The ancestry of Roda to the magnificent temples at Modhera, Sejakapur and Miani itself was a moral booster to the study of Roda group of temples.
THE SITE
Roda, as the word means in Gujarati, is in ruins and as mentioned earlier the site definitely had more temples than, which are extant. It may not be 109 as the local population claim, but the collection of sculptures in Baroda Museum definitely points towards an established religious site with temples dedicated to almost all in the Hindu pantheon. The ruining of the temples seems to be continuing as the site lost another temple within the span of 34 years [between 1926, (P.A.Inamdar's visit) and 1960 (U.P.Shah's stopover]. The closeness to the River Hathmati can be responsible for this, as even today after the monsoons a frequent visitor is forced to find a new way to reach the Temples. The sandy soiled topography no way helps the temples except the somewhat surfacing bedrock below the structures.

All the temples faces east showing a unanimity of the site planning. One is reminded of the other sacred sites as Mahakut and Pattadakkal [both in Bagalkot District, Karnataka] of the Chalukyans and the sacred structures on the Manmatha Tank at Hampi [Bellary District, Karnataka] of the Vijayanagara rulers. I have mentioned these sites for the specific reason that unanimity of cardinal axis did not mean the affiliation to a solitary patronage in both the centers. There is some other reason for this homogeneity. One, the builders were following the axial pattern already established by one of the structures, as in the case of Mahakut and Hampi where Mahakutesvara Temple and Virupaksha Temple [or at least the pre-Vijayanagara shrine to Pampa and Shiva] respectively determines the axis. Secondly, it determines by a Tirtha or a sacred water resource, which exist on the axis as in Pattadakkal, where River Malaprabha takes a curve. Roda seems to be following the pattern of Pattadakkal aligning the temples to River Hathmati, which again takes a picturesque loop[5].

There is a high possibility that the Shiva Temple [Temple 7] on the southern extreme of the site did determine the axial pattern and the rest followed it. In that case, the chronological placement of the Shiva Temple should be preceding the other structures. At this point, I would opt not to penetrate into the debate on chronology, but even in relation to axial placement, the Shiva Temple appears to be the earliest among the surviving structures. This can be understood thus; the curving of the river is prominent near the Shiva Temple, the temple was definitely connected once with the river through stone steps (now ruined) leading to the river and lastly except Temple 6 and 7 all the Temples are away from the river. Among these structures Temples No.3, 4 and 5 has their sacred water resource that is a Kunda on the east.

Affinity with the water is evident in the temples except those near the Kunda. Temple 7 has jutting out makara heads on the corners of the adhishtana. Temples 6 and 2 follow this. These do not prompt Shah to place them in one group. He was more concerned with the projections of the mandovara. Before we go further into the analysis, it will be pertinent to reiterate the grouping of the previous scholars as Shah and Dhaky. Shah clusters Temple 1,3,5 and 6 in one group which he consider as belonging to be earlier phase. Temple 7 is kept separate for its prominent mukha mandapa. Temple 2 for Shah was an extension of the Maitraka activities and thus an outlaw in the Roda group.

Shah’s theory reads thus, “A study of the recesses of the walls of the Roda shrines points that temples Nos. 1, 4 and 6 fall in one group, having, on each wall, only one central offset, while in another group represented by temples Nos. 3, 5 and 7 are two offsets on each wall and temple No. 2, representing the third group, has no offsets at all. In the second group, it appears that portico of No. 6 was a later addition while No. 7 has a big sabha mandapa not seen in other temples of the group. There is a figure of Lakulisa in centre on top of the doorframe, and above the doorframe, in niches, are figures of Brahma, Kubera, Siva, Ganesa and Vishnu. Temple No. 7 is possibly later than the other shrines.”[6]

The cause, which prompted Shah to group them, was the explicit difference in the mandovara [wall]. Temples 1,5 and 6 definitely followed a single pattern of the mandovara projection with the projecting bhadra and the supporting symmetrical projections on the sides which makes it three projections or tri ratha Mandovara. But the problem arises when he places Temple 3 along with the above mentioned temples. No architectural logic and scholarly intuition would support such a grouping. Another reason which prompted Shah to place temple 7 as the latest in the group probably was the size of Temple 7 which has a gudha mandapa and mukha chatushki to its credit. Shah follows the usual theory followed dealing with architecture, which is “the bigger the later”. He probably was not aware of the dated monuments of western India in the closer chronological proximity of Roda like Siva Temple at Kusuma [Rajasthan], dated 636 AD, had a developed gudha mandapa and mukha chatushki. On the other hand the temples which can be considered later to Roda as the Ranak Devi temple at Wadhwan, which almost clone Temple 3 at Roda, have no gudha mandapa.

Thus the question is when one tries to place the undated temples in the macro chronology of architecture development, especially in the case of Roda, it is the Temple 3 which has been the favorite of the architects of the consecutive years, not Temple 7. If Temple 7 was the most matured among the Roda group, why did not the later temples of Western India emulate it? Or how do we explain the sudden change of architectural pattern cutting down the gudha mandapa and reinstallation of it later? Shah, though concentrated on wall modulations and projection failed look beyond Roda and envisage the pratiratha concept being incepted in Roda 3 with the engaged columns. The next generation of temples from Gujarat especially Sourashtra as Ranak Devi Temple at Wadhwan, Muni Bawa Temple at Than, Siva Temple at Kerakot and Siva Temple at Kotai all follow the syntax of wall modulation in tune with Roda 3 than No.7.

As noted above the makara motifs on the corners of the adhishtana plays a great role in placing the temples in the micro chronology of Roda. If we group the temples into makara and non makara categories Temple Nos. 6, 7, 1 and 2 will come under the former and the rest of the temples near the kunda will be in the latter. There is a distinct stylistic affinity between the temples of these groups. Temple Nos. 7 and 6 are built on the River Hatimati with individual mounts and probably had separate access from the river. Temple Nos. 1 and 2 are built little away from the river, or appear so at present. As mentioned above the flooding or drying of the river would have been a major concern. The intention of symbolically keeping the temple afloat during the monsoons is clear with the makara motifs at the corners. The temples of the other group satisfy this water symbolism with Kunda in the front like Sankara's kamandalu would protect the temples by encompassing the floodwater in it.

Last but not the least point of disagreement with Dr.Shah is in regard to the doorway of the temples of the current discussion. While Temple 7 doorway exhibits austerity in decoration with merely Lakulisha figure on the lalata bimba, with no decoration on the jambs, the doorway of Temple 3 revives the tradition of magnificently decorated jambs with special addition of rupa sakha with cardinal deities. A similar doorway prior to Roda hails from distant Kosala region at Lakshman Temple at Sirpur, where the dasavatara panels grace the jambs. If this trend ended with Roda 3, one could have argued that the medievalization and the decadent severity were creeping into Roda region. The reality, however, remains that the temples, earlier to Roda by date, like the Maitraka and Saindhava temples, had austere doorways, as Temple 7 and the later temples of the region follows the doorway pattern of Temple 3.

The ambiguity, which engulfs the monuments of our purview, begins from their historical placement itself. Though scholars like Shah and Dhaky place them around Eighth century AD on the stylistic grounds, they are not sure who the people responsible for this development. This mystification arises for the lack of documentary evidence in any form from the region. As the area shifted from one power to the other it was difficult pin point the patron the temples of our concern. Dhaky cites the example of Kiradu temples to illustrate such chaotic chronological or dynastic affiliation thus,“Take the example of the temples of Kiratakuta [Kiradu]. Who the authors of these temples were, is still not certain. Agreed, Cahamanas as well as the Paramaras occupied this city in the later half of twelfth century; but the temples in question were already in existence there, the latest being the being older by three generations than the known facts of history. Kiratakupa temples suggest stylistic affiliation with both upper Rajasthan as well as lower Rajasthan, even Gujarat. What dynastic label shall we attach to them?”[7]

The sudden change of architectural syntax can not be considered as a habitual development. As Dhaky and Shah observes if we keep Roda in mid Eighth century the credit of building the site should go either to the Rashtrakutas or the Pratiharas. The dynastic affiliation should come either from south or north. As mentioned earlier Roda shows affinity to Deccan region. This hypothesis can be emphasized further with two more structures from south India. They are Sangamesvara temples at Mahakuta and Kashivisvesvara at Pattadakkal both regarded generally as Chalukyan structures. Except Mallikarjuna and Virupaksha temple no other structures in Pattadakkal has dedicatory inscriptions thus making all the more difficult to place monuments chronologically. Dantidurga the Rashtrakuta ruler was a feudatory of the Chalukyas in the middle of the 8th century. Their existence in Gujarat can be traced right from 738 AD when Dantidurga joined hands with Pulakesi of Navasari branch of Chalukyans to push Junnayid’s invasion of Gujarat in 740 AD. Dasavatara cave inscription at Ellora suggest that Dantidurga captured Malawa and other countries which further confirmed by the Sanjan plates of Amoghavarsh stating Dantidurga made Gurjara lords and other kings as door keepers [Pratiharas] at the Hiranya Garbha ceremony at Ujjain. The Rashtrakuta inscriptions from Sanjan [Surat District], Kavi [Bharuch District], Akota [Baroda District], Kapadvanj [Nadiad District] in Gujarat shows their possible activity in Gujarat. More over there are clear indications of Rashtakuta architectural activity in Malwa which evident through the monuments at Vasavi and to an extant Dhamnar. Indragarh inscription of 767 AD by Nanappa claims his lineage to the Rashtrakutas. Thus when the areas around Roda regions have evidences supporting the Rashtrakuta activity, there is no need to write off the specific areas to the Pratiharas. The historical data further refers to the Pratihara-Rashtrakuta clashes as it informs of the Rashtrakuta king Govinda III’s victory over Pratihara king Nagabhata II in the battle at Bundelkhand.

If we accept the Rashtrakuta hegemony over Gujarat, we may have to find out the Rashtrakuta architecture in Gujarat. It will not logical to argue that the Rashtrakutas never built in Gujarat, as they have been the most voracious builders of the early medieval period. They try to spot their hegemony with a monument whether it is in Ellora, Hampi or in Malwa as medals of war or as Castriota states “advertisements of political suzernity”. Distinctness invites attention and attraction which the watchword of present day advertisements. Rashtrakutas probably addressed this aspect by building unlikely structures at odd places. In Ellora where all the caves were following the horizontal carving, they have hewn out monolith structures. In Hampi they built temple in sandstone just to contrast the granite and schist structure. In Vasavi they attempted to repeat the Ellora monolith, in Dhamnar they carved the first Nagara monolith.

The same feat they must have experimented in Roda. The situation prior to Roda had drastically changed after the construction Temple No.3. No stapati was willing to return to the old Maitraka Saindhava style of masonic buildings because latina type of structure as Temple 3 have promised greater possibilities of vertical and horizontal expansion. Temple 3 had just one anuratha in the form of the engaged column duly decorated. Ranak Devi temple merely follows it while Muni Bawa temple introduces to anurathas with figural decoration on them. At Kerakot and Kotai the madovara become more vibrant with multi facets.

The issues remain to be addressed are what exactly is the Rashtrakuta style of architecture and will the shrines at Roda fit into it. It will be difficult task for architecture historian to tackle these as the only inscribed Rashtrkuta monument is Cave 15 [Dasavatara] at Ellora. This was inscribed soon after Dantidurga claimed his independence from his Chalukyan over lords. In Kailashnath at Ellora we do not have site inscription but copper plate which proclaim that King Krishna was the reason behind the magnificent structure. As the Rashtrakuta Empire extended from Malwa in north to Tondainadu in south they came into contact with already mature styles of Indian temple architecture, that is Nagara, Dravida and Vesara. As many a dynasties stuck one of these and accepted the specific style as their characteristic, Rashtrakutas shuffled them and placed them in the contrasting regions. Thus they need not have to customize the style.

The political hegemony can be expressed through three ways. [1], the rejection of extant style and introducing a contrasting style, which would make the people, forget the older style and their creators. [2], by following the same style and claiming the political legitimacy. [3], through introducing new elements in the existing style and customizing it. The selection of the path depends upon the process of the transfer of power and the definition of the relation between the rulers and the ruled.

Many a rulers in India have followed the second method, as the political legitimacy was crucial to have the acceptance of the subjects. Moreover many of these rulers were the subordinates of the previous dynasties at times having even blood relations. Ucchakalpas of Nachna conformed this method following the Gupta Style. Cholas to an extant followed the Pallava idiom precisely for this purpose. The examples of third method can be observed in the colonial British architecture in India. Rashtrakutas probably followed the first and second method. In Chalukyan regions or even in Ellora they followed the Chalukyan style with the technical collaboration of the Pallava carvers. The regions away from these areas they followed the first method. The reason why the Solanki rulers following Roda was to reclaim the political legitimacy of Chalukyan on Gujarat through Rashtrakutas.

CONCLUSION
Apart from being a group of magnificent monuments, which has drastically changed the architecture pattern of Western India, Roda group of temples appears as a statement made by the architect or the patron with a specific intention behind. The present scholar is suggesting a Rashtrakuta lineage to the temples because of their affinity to the Deccan idiom. Thinking along the line one has to justify the Rashtrakuta hypothesis. The problem arises when we realize that there are no dated authentic Rashtrakuta structures except the Dasavatara Cave Nandi mandapa at Ellora. It become more complex if one attributes the structures to the Pratihara dynasty, as there will be no justification for why there is shift in the style between Osia and Roda. Moreover the Kashivisvera temple at Pattadakkal which shares an emphatic consanguinity to Roda 3, seems to be a Rashtrakuta structure for its placement and style.

This hypothesis is based on the seven extant temples at Roda and their sculpture. The temple building activity at Roda probably has continued at least for two centuries as the Baroda Museum sculptures and the 13th century step well at the nearby village would suggest. In the process of the clashes between the Pratiharas and Rashtrakutas Roda probably might have changed hands but except the minor stylistic shift between the makara group and non-makara group. Along with the stylistic analysis the political interpretation was necessary, as the monuments have changed the visual sense of the people of the region. The complexity of medieval, which will be mainly playing with space and the intricacy of wall modulation starts here at Roda. As architecture history has been perceived as study of terminology and parched description it was inevitable to look beyond the stylistic contemplation. As David Castriota notes, “From earliest times works visual art have directly served and reflected the beliefs and institutions of the societies that produced them. Often careful interpretation of works of art can yield a wealth of cultural in formation that might otherwise lost.”

The temples have been crumbling down since centuries. The local farmer would not even bother to look at these illustrious structures. The zealous master builder has created a history here. The resolute benefactor had his own enigmatic motivations. The current study attempts to delve into these ere they are completely disseminated.
* * *

[1] UP Shah, 1960, pp.88
[2] PA Inamdar, Some Archeological Finds in the Idar State, 1926.
[3] PA Inamdar, Some Archeological Finds in the Idar State 1926.
[4] UP Shah, 1960, pp. 114
6 It may be curious to note that the only imperial house, who had constructed in the above-mentioned sites, leaving Roda for the time being, is the Rashtrakutas.
[6] UP Shah, 1960, pp.
[7] MA Dhaky, 1968, pp. 309